or S 2 ·− were detected). This was also the only radical species observed by Barchasz et al. [ 12 ] upon discharge of a sulfur electrode in tetraethylene glycol dimethyl ether (TEGDME). None of these studies, however, gives a quantitative estimate of the free radical content in the electrolyte, as compared with the various S n 2− dianions. Is the S 3 ·− radical prominent upon cycling in Li–S cells using dimethoxyethane (DME) and 1,3-dioxolane (DOL) solvents? In this event, could its reactivity be held responsible—through electrolyte decomposition—for capacity fading over extended cycling? If not, could the practical capacity of a Li–S cell be augmented by favoring free radicals by tuning the dielectric characteristics of the electrolyte? Herein, we assess the effect of sulfur radical species formed upon cycling of Li–S cells; in particular S 3 ·− . Based on our unequivocal observation of sulfur radicals in a Li–S cell by X-ray absorption near edge structure (XANES)—for the fi rst time under operating conditions—using an EPD solvent (DMA), we show that radicals are not stabilized in glyme-based electrolytes. However, we do show that S 3 ·− reacts with DOL at elevated temperatures, while DME remains intact. In contrast, the much greater dissociation of the anion precursor, S 6 2− , to the trisulfur radical in donor solvents such as DMA and DMSO—where trisulfur is in high concentration but nonreactive—surprisingly and importantly allows the full utilization of both sulfur and Li 2 S. The effective solvation of the latter results in the complete absence of an overpotential on charge. Chemical incompatibility between the lithium metal negative electrode and EPDsolvents can be overcome with anode protection, demonstrating their applicability as electrolytes for the Li–S or Li 2 S batteries in hybrid cells. The nature of sulfur species in nonaqueous solutions of alkali polysulfi des has been well established since the 1970s, using UV–Visible or electron spin resonance spectroscopies in various solvents. [ 9,10,12 ] It is not the scope of this work to repeat such experiments, and we accept that solvents with low donor number (DN), such as tetrahydrofuran (THF) contain exclusively dianions, and in particular S 4 2− and S 6 2− that give rise to a yellow coloration ( Figure 1 a,e). In contrast, solvents with high DN give intensely colored blue solutions owing to the S 3 ·−
Read full abstract